If you think keeping up with the content tsunami is hard in English, try it in multiple languages. Ensuring an omnichannel experience across all countries means translating everything: Tweets, blogs, white papers, websites - if it's part of the buyer's journey, it needs to be in a language the customer understands. If it isn't, they won't buy.
In research report "research called Can't Read, Won't Buy," market analysis company Common Sense Advisory surveyed online shoppers from 10 non-English speaking countries. Sixty percent said they never or rarely buy from sites not in their language.
Translation Services Cost
Translation may drive sales, but done correctly, it's rarely cheap. And the high volume of content that omnichannel requires can quickly deplete a marketing budget unless you know where to save.
And for those who are considering free tools like Google Translate to keep down costs, they need to think again.
Related Story: A 5-Point Checklist for Finding a Translation Management System
Free Translation Tools Make Mistakes
According to Rick Antezana, president of the Association of Language Companies (ALC) - the national trade organization for translation businesses, "Marketing content is one of the areas of professional communication that is not really ever well-suited to applying free online tools for translation." Machine translation (MT) - the industry term for free and paid translation programs - "is incredibly well developed, but never perfect," he explains. "The errors MT sometimes introduces can be damaging to brands when they are comically wrong or sometimes even offensive." Like "Mondelez," for example. It's Kraft's international brand name, but in Russian means something completely different.
That's not a mistake any marketer wants to make and is a problem a human translator could have caught. But, as Antezana adds, "The main disadvantages of human translation are that it's not free and it takes more time than machine translation." So thinking back to budget, how do you know when to pay the professionals and when to use free, online tools?
Allison Ferch, executive director at the Globalization and Localization Association - ALC's international counterpart - says it's complicated: "We could say only use Google Translate for marketing materials if you don't mind damaging your brand or losing customers, but of course it's not that simple." In the translation battle between machine and human, she says consider audience, context, medium, and risk: "If getting it wrong would spell disaster for your brand, then go with professional translation and localization."
Learning Opportunities
Arle Lommel, a senior analyst at Common Sense Advisory, agrees: "Content falls on a continuum from low-value materials produced by third parties, such as user reviews that may require little or no human attention, all the way up to core marketing concepts," like your brand name or slogan. Higher-level copy, he explains, is definitely not choice for free tools: "Think of it this way: You may spend hundreds of thousands of dollars to craft your message in English, but using a free service is throwing that investment away for other languages." Investment is the key, though - specifically, return on it.
Related Story: Why Machine Translation Matters in the Modern Era
Consider a Hybrid Approach
No marketer wants bad quality, but pragmatically some budgets can't be expanded. In the balance, there is a hybrid option, machine translation post-editing (MT+PE). Lommel describes it as "a human 'post-editor' [who] check[s] the output from a machine." But these professional post-editors don't use Google. They review their own, higher-grade machine translation. "Professional machine translation is adapted and tuned to your own content," Lommel explains, but free online tools mix everyone's translations together. The result, he adds, could be "translations using your competitors' terms and language, mistranslations that arise because you use specialized language, and untranslated content that slips through."
Antezana says this hybrid option is best for high volume, low risk, and/or low importance content - like user support or help files. "The fewer eyeballs on potential content for translation and the higher the volume, the more appropriate it would be for machine translation," he adds.
Both Antezana and Lommel say MT+PE is best for quick turnaround projects that would take a human translator longer to complete solo. And Lommel contends this option works best when the original and new languages are more closely related: "If you are translating between English and German or French - or between French and Spanish, you can expect better results than if you are going between English and Chinese or Japanese."
In any language, Lommel adds, it's important to remember you don't have to pick just one option. Just as MT+PE combines machine and human translation, you can mix and match the translation processes you use. "It isn't that you have just a choice between machine translation and human translation, but rather that you can adapt processes to meet your needs and to balance outcomes against your time and budget," he explains.